Does anyone have any indications of possible CPU savings made by CI size or logger changes? Not really noticed SMF address space itself as being a particular CPU hog, or would savings be seen in all address spaces that write out SMF?
We're pushing out over a billion records a day. On 10 October 2012 17:18, Robert A. Rosenberg <[email protected]> wrote: > At 06:29 -0700 on 10/10/2012, Lizette Koehler wrote about Re: Change MANx > CISIZE on the fly: > > > Of course if that worked, you would need to dump all of your MANx files >> before the cycle and accept some loss of SMF data. >> > > You can allocate new MAN datasets and point at them at IPL time. Then just > manually run the dumps again the old (small-CI) datasets before deleting > them. > > > ------------------------------**------------------------------**---------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
